1

The Exchange Control System under Apartheid

Roy Havemann and Lusanda Fani1

Prepared for

“The Economics of Apartheid: An empirical investigation into its causes and

consequences”.

* First draft *

* Version 11 February

Abstract

Exchange controls were part of a complex system of maintaining some financial

stability during apartheid, particularly as the system began to collapse due to

sanctions and deteriorating external and internal conditions. The centrepiece of

the system was a dual exchange rate system, introduced almost by accident in

1961 to protect against possible negative consequences after the country left the

Commonwealth. From time to time, the government tried exchange control

liberalisation, particularly following the recommendations of the De Kock

Commission. The economic deterioration of the early 1980s, however, and the

Rubicon speech of 1985 changed everything. The dual exchange system was re-

introduced, and restrictions on capital expatriation were substantially tightened.

This protected the economy, to some extent, from sanctions and disinvestment.

However, with this short term benefit came with medium- to long-term damage.

Capital was trapped inside the country, and the industrial structure became more

1 The authors would like to thank James Cross, Raymond Paulo, Jonathan Leape, and Lynne

Thomas for patiently explaining many aspects of the exchange control system, and to the Centre

for Research in Economics and Finance in Southern Africa at the London School of Economics

for access to their library. This paper is written in the authors‟ personal capacity, and does not do

not necessarily represent the views of the National Treasury or the Reserve Bank.

2

and more concentrated – by 1985, the four largest corporations controlled

80 per cent of the Johannesburg Stock Exchange. (This rose to 81 per cent by

1990). In addition, in manufacturing a quarter of production was controlled by

four firms. These firms could not expand offshore and expanded into all manner

of non-core activities – South African Breweries, for example, controlled not only

the local beer market, but had substantial interests in hotels, gaming and other

leisure areas. By making it difficult for investors to repatriate capital, foreign

investment was discouraged. And, the centralised system of exchange rate

management meant that a functioning market-determined forward market was

stunted. The era shows how capital controls can create substantial distortions,

many of which persisted even after the liberalisation of the regime. Indeed some

of these distortions remain today (and some of the controls too). This highlights

the need for ongoing reforms to make the economy more competitive and more

integrated, and develop market-based systems and „macroprudential‟ rules to

manage exchange rate and other external risks.

3

Table of Contents The Exchange Control System under Apartheid .................................................... 1

Abstract ................................................................................................................... 1

1 Introduction ..................................................................................................... 4

2 Exchange control, 1961 – 1983 ...................................................................... 4

2.1 The legal background – Currency and Exchanges Act of 1933 ............... 4

2.2 The events of 1960 ................................................................................... 5

2.3 The dual exchange rate system................................................................. 6

2.3.1 Dual exchange rates in theory ........................................................... 6

2.3.2 The blocked rand –1961 to 1976 ...................................................... 7

2.3.3 The securities rand – 1976 to 1979 ................................................... 8

3 The Rubicon, and re-introduction of exchange controls ................................. 9

3.1 The years before the Rubicon speech ....................................................... 9

3.2 The Rubicon speech, debt standstill and exchange controls .................. 11

3.3 The new financial rand system ............................................................... 17

3.4 Capital flight ........................................................................................... 18

4 Consequences ................................................................................................ 19

5 Conclusion .................................................................................................... 22

6 Bibliography ................................................................................................. 24

4

1 Introduction

The 1963 – 1972 period has been described as the „golden age of apartheid for

those class forces that benefited from the system‟, with rapid economic growth

and rising standards of living for all sections of the economy, but particularly for

the protected white group.2 The 1985 period onwards was in many ways the

complete opposite, and all sectors of the population suffered as the system turned

inward and autarkic, and the apartheid system became increasingly unsustainable

both morally and economically.

This paper outlines the exchange controls that were used during the apartheid

period, in particular explaining the economic conditions that led to their use. It is

arranged as follows: section 2 briefly discusses the exchange rate arrangements

prior to 1985, focussing on the dual exchange rate system; section 3 spends some

time on developments following the August 1985 Rubicon speech, including the

introduction of multiple exchange rates and capital flow restrictions. Section 4

considers possible consequences, highlighting how the system influenced

industrial structure, investment and savings patterns both before and after the

democratic transition. Section 5 concludes and points to some lessons that the

period holds for today.

2 Exchange control, 1961 – 1983

2.1 The legal background – Currency and Exchanges Act of 1933

On 28 December 1932, the Union of South Africa left the gold standard, taking

the country (and indeed much of the world) into a new monetary policy era. The

first instinct was to allow currency to float (De Kock, 1954). However, South

Africa was running a substantial current account surplus due to gold exports, and

by the end of January 1933, the South African Reserve Bank had decided to buy

2 Davies et al (1984), quoted by Moll (1990)

5

up additional sterling and maintain a fixed exchange rate system. The legislative

framework for currency policy was then introduced in the form Currency and

Exchanges Act 9 of 1933, and accompanying regulations. The Currency and

Exchanges Act 9 of 1933 ("the Act") regulates legal tender, currency, exchanges

and banking. Section 9 still empowers the President to 'make regulations in regard

to any matter directly or indirectly relating to or affecting or having any bearing

upon currency, banking or exchanges'. The currency of the South Africa is the

South African rand and other legal tenders can be certain gold coins, including the

Krugerrand.

This Act is indeed still in force at the time of writing (2013), and provides the

legal framework for the entire exchange control system. However, when it was

passed, it was not intended to support the draconian measures that needed to be

taken thirty years later.

The system worked reasonably well for a sustained period of time, bringing

relative stability to the rand during the Second World War, the post-war

reconstruction and the global boom period that followed.

2.2 The events of 1960

From 1960, everything changed. On 3 February 1960, British Prime Minsiter

Harold Macmillan addressed the South African Parliament, saying that a “wind of

change … [is] blowing through this continent”. The following month, on 21

March 1960, a demonstration against pass laws was held outside a police station

in Sharpeville. Accounts vary as to what and how the situation deteriorated3, but

the outcome was that police opened fire on the crowd, killing 69 people. Under

severe international pressure, and following a domestic referendum, South Africa

left the Commonwealth in 1961.

3 A detailed account is contained in the Truth and Reconciliation Commission report, chapter 6.

6

The authorities were concerned about the implications for the economy of an

outflow of capital, and the economic stresses that could arise, and immediately

substantial steps were taken to tighten exchange controls.

Using the framework of the 1933 Act, two proposals were accepted: first, a dual

exchange rate system was introduced, and second administrative controls on

currency conversion were substantially tightened (Kahn, 1991). The Exchange

Control Regulations, 1961 was promulgated by Government on 1 December 1961

and amended in 1 November 2010. Together, these two measures became known

as “exchange controls”, and today many aspects of the latter intervention remain.

Each of these measures is discussed in more detail below.

2.3 The dual exchange rate system

2.3.1 Dual exchange rates in theory

Dual exchange rates4 have had periods of popularity in many emerging markets

(Kiguel & O'Connell, 1995). Theoretically these are very elegant systems to deal

with balance of payments volatility: typically, current account transactions take

place at the official exchange rate, and capital account transactions at a market

determined exchange rate (Flood & Marion, 1982). Thus international reserves

are unaffected by capital outflows (which lead instead to a depreciation of the

parallel rate). Relative to a unified float, the effect of a dual system on domestic

prices is limited because all current account transactions take place at the official

(pegged or managed) exchange rate. Even Italy and France briefly had dual

systems in the early 1970s; to guard against temporary external shocks brought

about by the collapse of the Bretton Woods fixed exchange rate system (Kiguel

& O'Connell, 1995).

Indeed by 1992, thirty-three countries operated some form of non-unitary

exchange rate systems, albeit systems that varied quite substantially between

4 Both legal systems and black-market systems.

7

countries and were quite different to the financial rand system used by South

Africa (Garner, 1994).

2.3.2 The blocked rand –1961 to 1976

It appears the intention was not for there to be a formal dual exchange rate system

– indeed the view of the authorities was that they were simply blocking funds of

non-residents5, and that this block would only be a temporary measure to deal

with the immediate crisis (Farrell & Todani, 2004). However, it was obviously

simple to derive the shadow exchange rate during this period, based on arbitrage

conditions, and thus the emergence of the dual exchange rate (Barr & Kantor,

1983).

In practice, the dual exchange rate system entailed the introduction of a “blocked

rand” for “non-residents” and a “commercial rand” for residents (Garner, 1994).

The dual exchange rate system could be seen as the centrepiece of the exchange

control system. Firstly, it introduced the concepts of “exchange control residents”

and “exchange control non-residents”, and differing treatment between the two.

The intention was partly to penalise non-residents for selling South African

assets; partly to support a relatively stable commercial rand, and in a related

objective, guard against the inflationary impact of a rapidly depreciating currency.

The initial “blocked rand” system operated from June 1961 to February 1976. In

this system, the rand proceeds from non-residents‟ sales of listed securities and

capital assets were designated as “blocked”. Funds could be used to buy quoted

South African securities, which could then be transferred to London and sold. In

5 Technically, over half a century, the blocked rand system remains – South Africans who have

emigrated for exchange control purposes, i.e. changed from excon residents to non-residents,

must retain assets in a blocked rand account. Until 2010, exiting these funds required the payment

of a 10 per cent levy (see National Treasury 2010b).

8

practice, a discount was created between these securities that were freely

transferable, and those that were not. This was for two reasons: first, non-

residents were net sellers of securities as they were trying to exit the country;

second, residents were substantial buyers of securities as they had few other

investment options. The implication was that the rands in a blocked non-resident

account were worth less compared to a rands in a resident account.

2.3.3 The securities rand – 1976 to 1979

The 1970 Franszen Commission of Enquiry into Fiscal and Monetary Policy in

South Africa6 argued to retain exchange controls, in view most importantly of

political factors that could cause capital flight (Kahn, 1991). Nevertheless, the

system needed improvement. In February 1976, the blocked rand system was

altered to a “securities rand system”. The new form allowed direct transfer

between non-residents – i.e. non-residents could exchange securities rands with

one another. Before March 1978, a rule allowed the sale proceeds of government

and parastatal bonds purchased with securities rands to be freely transferred

offshore. This was abolished. The effect was to reduce the incentives for non-

residents to invest in such stocks.

At the same time, political pressure began to grow in the rest of the world for

sanctions, and by the mid-1970s, a full scale economic boycott of South Africa

was increasingly likely (Spandau, 1979). Sanctions began to bite in 1974, when

Switzerland imposed a cap on borrowings to South Africa (Ovenden & Cole,

1989), and the first evidence of sanctions was the decision by Midland Bank in

1975 to stop lending to the South African government or its agencies, and to

completely stop any activity with South African counterparties by 1978. After the

events of 1976, Dutch commercial banks agreed to end lending to the

Government, and export finance restrictions followed in 1978. More ominously,

from 1977 loans from a number of banks in the United States were terminated. In

6 More details on the commission are available in Kahn (1991), Garner (1994) or Leape (1991).

9

the twelve months to July 1985, seven US states and twenty-five US cities acted

to move their business away from companies with South African exposure.

3 The Rubicon, and re-introduction of exchange controls

3.1 The years before the Rubicon speech

The 1980 to 1984 was particularly difficult for South Africa, as economic

conditions deteriorated significantly. The commercial rand devalued substantially

between 1980, with the devaluation gaining momentum in the second half of 1984

Notwithstanding the depreciation, the export-led growth that characterised the

1973 to 1979 period came to an end, and external trade grew slower than GDP

(Jones, 1994). The deterioration of the balance of payments was accompanied by

a substantial fiscal deterioration: Between 1980 and the end of 1984, South

Africa‟s total external debt rose from 20.3 per cent to 45.7 per cent of GDP. The

government found it increasingly difficult to raise long-term foreign currency

borrowings, as many international banks began to feel political pressure to reduce

their exposure to South Africa (Ovenden & Cole, 1989). External debt both

increased in size and deteriorated in quality and become more short-dated as the

state struggled to raise long-term financing. Over the same period, the proportion

of debt maturing within twelve months rose to 65 per cent, up from 52 per cent. In

contrast, private sector borrowings rose substantially, particularly borrowing from

abroad. In part, this was due to extraordinarily high interest rates, which touched

10 per cent in real terms in 1984. At the same time, consumer price inflation

touched 15.3 per cent in 1981.

To make matters worse, the gold price slid from a peak of US$612 in 1980 to

US317 in 1985, and gold as a share of export revenues fell from 51 per cent to

44 per cent (Leape, 1991).

10

The sanctions detailed above did not helpi. By mid-1985, the Apartheid state was

faced with a wholly unsustainable fiscal position, lack of access to international

capital markets and a current account in difficulty. The final straw was the

decision by Chase Manhattan Bank to stop rolling over loans to South African at

the end of July 1985 (Waldmeir, 1999). An economic crisis became a reality.

Figure 1: Financial rand and commercial rand from 1985-1995

11

3.2 The Rubicon speech, debt standstill and exchange controls

On 15 August 1985, on the opening of the Natal National Party Conference in the

Durban City Hall, then President P.W. Botha gave a speech on the future of the

apartheid state7. The “Rubicon speech” was widely hyped ahead of its delivery –

indeed it was expected that it would signal how the Apartheid state would slowly

be dismantled (Waldmeir, 1999). The economic circumstances, after all, were

very difficult. Indeed, the speech did weakly promise some areas of reform,

hinting even at universal franchise. But the tone was highly defensive, and the

underlying message was that the Apartheid state would remain intact.

As the balance of payments rapidly deteriorated, increasingly desperate measures

were taken to maintain economic stability, essentially economic autarky.

Over the course of August 1985, the (still unified) rand fell from R2.18 to R2.77,

a depreciation of 27.3 per cent. The following major steps were taken in a very

short space of time to deal with the economic crisis that followed the speech:

On 31 August 1985 (just two weeks later), the government announced a

debt standstill. The standstill stipulated that capital payments on foreign

debt would be suspended until an agreement could be reached with

creditors, although interest payments on debt would continue

uninterrupted (Leape, 1991). The debt covered by the standstill totalled

US$13.83 billion, about 57.5 per cent of the total outstanding8.

The following Monday, 2 September 1985, the dual exchange rate system

was reintroduced. This had the effect of appreciating the commercial rate,

which rose 25 per cent that day alone, returning to R2.22, or

approximately the same level as at the beginning of August. It averaged

R2.49 for the month of September. The new financial rand, however,

7 The full speech is available at http://whatsimportantinlife.wordpress.com/2012/11/09/the-real-

rubicon-speech/ 8 R50.76 billion in 1985 rands; or R124.47 billion in 2013 rands. To put this in perspective, the

debt stock in February 2013 was RXX billion.

12

averaged R2.89 per US$ during August compared to an average rate of

R2.49 for the month (i.e. the discount in the first month was

15.8 per cent);

Exchange controls were tightened even further to protect capital flight;.

Summary of 1985 manual:

Non-residents: The local sale or redemption proceeds of non-residents

owned South African assets were not allowed to be converted into foreign

currency at the commercial rand but had to be retained in South Africa

with Authorised dealers in the form of Financial Rand balances. The

financial rand balance was transferable between non-residents and could

be reinvested in South African quoted securities and other investments

approved.

Capital controls restricted non-residents and in turn affected inward

investment into South Africa. In terms of foreign direct investments, local

borrowing by foreign-owned corporates9 was limited to 300 per cent of the

foreign capital invested in South Africa. Wholly-owned local enterprises

by a foreign firm were also allowed a 300 per cent local borrowing limit

on shareholding capital. The 3:1 ratio was applicable and local financing

above required approval from the South African Reserve Bank (Financial

Surveillance Department).

Residents: Residents must not buy, borrow or sell foreign currency except

an Authorised Dealers with permission granted by the Treasury. Private

individuals were allowed to invest in foreign assets but the allowance for

outward investment was set at R200 000, subject to tax obligations. Assets

in excess of the emigrant allowance were held in blocked Rand accounts

South African institutional investors (long-term insurers, pension funds

and unit trusts) were only allowed to enter into asset swaps of 5 per cent of

their total assets.

9 Applies to foreign-ownership of 75% but the greater degree of the local participation relaxes the

borrowing limits.

13

In addition, the authorities began intervening heavily in the commercial

rand market to bring some stability, at least to the domestic currency

market, which ensured that the rand averaged R2.60 to the dollar for the

remaining of 1985; and

Interest rates were raised by 7 percentage points from 18 per cent to

25 per cent.

Increased reliance on forward cover provided by the Reserve Bank, to the

extent that cumulative forward cover losses reached R15 billion by 1989

(Ramos, 1991). This is the equivalent of R133.5billion in 2013 rands10

.

In all, the appreciation of the current and rise in interest rates probably had the

immediate effect of preventing further inflation, but the more important effect was

to impose a stringent balance of payments constraint on the economy. Capital

flight became increasingly desperate, and the nature of the exchange control rules

of the time highlighted how South African exchange control residents were

restricted from almost any attempt to exit cash or assets, including strict rules on

travelling with cash, rules on travelling with cash, and things such as rules on

yachts, and amount of gold jewellery that a person could travel with.

Notwithstanding a new draconian order, the exchange rate was quite volatile. On

2 September 1985 (the day the standstill was announced), the rand fell from

R2.77 to R2.22, falling 24.8% against the US dollar.

10

Ramos (1991) provides a detailed analysis of the use of forward cover during this period. In the

absence of a fully-fledged market-determined forward cover market, the Reserve Bank quoted and

guaranteed forward rates, and thus took on the risk of adverse movements.

14

Figure 2 Commercial and financial rand, 1980 to 1994

The combined effect of the Rubicon speech and sanctions on the ability of South

Africa to borrow is most dramatically illustrated by Table 2. International capital

borrowings were zero between 1986 and 1987, and only began to improve slightly

from 1988 onward. Long-term syndicated loans were most common, but later

investors went into the securities market and private domestic firms and public

corporate became dependent on credit which South African banks raised from

abroad (see table 3 and 4). In terms of table 3, the composition of the South Africa

debt substantially changed. Firstly, the proportion of the short-term loans rose

rapidly from 49 per cent to 72.0 per cent. The second change was in the

composition of debt where there was a movement from conventional long-term

and medium-term loans to securitised borrowing into bonds

15

Table 1 South Africa’s real borrowings on the international capital markets,

1980 to 1989 (1982 US$ millions)

Public

bonds

Private

bonds

Syndicated bank

loans

Total gross debt inflows

1980 198 228 424 849

1981 0 97 409 507

1982 163 164 1485 1812

1983 250 221 785 1256

1984 550 465 432 1447

1985 464 301 0 765

1986 0 0 0 0

1987 0 0 0 0

1988 0 31 0 31

1989 0 105 0 105

(Garner & Leape, 1991)

16

Table 2 Foreign debt in South Africa from 1980 to 1986

For debt

US$ million

For debt

Rand

million

Foreign debt

as % GDP

Short-term

loan for

debt

Short-term loan

debt as % of total

debt

1980 16890 12595 20,3 6133 49.1

1981 18889 18081 25.5 10573 58.5

1982 22609 24249 30.4 13725 56.5

1983 23294 29116 32.5 19169 65.8

1984 24294 48230 45.7 32804 68.0

1985 23473 60142 50.0 43273 72.0

1986 22593 49513 35.2 35816 72.3

Source: South African Reserve Bank (SARB)

Table 3 Percentage structure of South Africa’s foreign liabilities

1970 1980 1983 1984

Equity capital incl.. shares, reserves 60.7 48.6 34.7 27.2

Loans to private sector 23.6 22.6 29.1 37.1

Loans to public corporations 5.2 16.5 13.7 11.5

Loans to central government 11.1 13.7 22.7 24.3

Source: SARB

17

The binding nature of the balance of payments constraint is most dramatically

illustrated by the balance of payments statistics for the period. The discussion

above highlighted that the persistent current account deficit was not adequately

covered by inflows on the financial account.

Table 4 The South African balance of payments, 1982 – 1990

(Current Rand Millions)

Year Exports Imports Current

account

Capital

movements

Gross

reserves

Estimated

net

reserves

1982 22,220 26,042 -3,696 +4,748 4,318 2,090

1983 23,617 24,214 -451 +158 4,968 1,485

1984 29,004 31,660 -2,602 +1,236 5,323 298

1985 41,394 36,348 +5,087 -8,321 5,889 -1,030

1986 47,420 41,371 +6,114 -5,072 5,725 527

1987 50,203 44,171 +5,995 -2,851 7,940 4,672

1988 58,575 55,932 +2,728 -6,208 6,705 1,609

1989 68,067 65,179 +3,108 -4,345 6,904 9

Source: Leape (1991)

The table highlights the severe pressure the balance of payments constraint

imposed. Reserves were low by the time of 198511

. By the end of 1984, South

Africa gross reserves (holdings of gold and foreign currency) equalled

R5.3 billion. Net reserves (gross reserves less foreign liabilities held against

reserves) were all of R298 million, or a full 3 days of imports (Leape, 1991).

3.3 The new financial rand system

In contrast with the 1961 arrangement, the intention with the 1985 financial rand

was specifically to create two currencies. There were two tiers, a commercial rand

and a financial rand. Any applications to move money in or out of South Africa

11

See figure below in Appendix 1

18

will pass through one of the two tiers. The commercial rand: rate for trading

commodities in South Africa, and all imports and exports paid with this rand in

addition to a number of other transactions (e.g. dividends, royalties and licence

fees). In addition, the commercial rand was a dirty float.

One important component was to manage foreign exchange reserves through gold

sales. It was required that the entire output of the gold-mining industry had to be

sold to the SARB within one month of production. The SARB played the role of

agent for all international gold sales, gold swaps, and internationally currency

operations entailing gold. A complex system of export and import permits for

many goods, administered jointly by Reserve Bank and Department of Trade and

Industry existed.

The financial rand, applied to four categories

Investments in South Africa by non-residents, covering purchases of equities,

bonds etc. (Loans go through commercial rand system).

Capital imported by immigrants, i.e. those moving to South Africa

Proceeds of the sale of South African securities when they are repatriated by non-

residents, i.e. conversation rate that applies is the financial rand conversion rate

Capital transfers from South Africa to foreign destinations

3.4 Capital flight

Though extraordinarily onerous, the regime could not halt continued capital flight,

although the more onerous regime after 1985 assisted in reducing most flight

(although this may have been because the exchange control system was widely

flouted12

. Estimates on the extent of capital flight vary depending on the type of

methodology used. Using a balance of payments approach, about US$12.4 billion

12

The Minister of Finance announced two Voluntary Disclosure Programmes, one in 2004 and

related Tax Amnesty Process and one in 2010. The total led to R68.9billion being declared of total

assets from 43681 applicants.

19

over the 1970 to 1988 period, and in addition trade misinvoicing accounted for

another (Kahn, 1991).

Figure 3 Capital flight, % GDP (1979-1988)

Source: Adapted from Kahn (1991: Annex B). The author details the methodology used in the

different approaches, noting that the BOP capital flow and Trade misinvoicing estimates are not

alternatives estimates, but rather complementary, and should be added together.

4 Consequences

Many of the effects of exchange controls lingered beyond the end of apartheid,

particularly the autarky brought upon by the debt standstill. Indeed, some of the

controls remain, and understanding both intended and unintended consequences

of the system is thus important. The following possible consequences have been

identified:

20

Reduced foreign investment. By making it the repatriation of funds both

difficult and expensive for non-resident investors, exchange controls almost

certainly deterred foreign direct investment (Farrell & Todani, 2004). Investors

into South Africa had to buy South African assets at the commercial rand rate, but

could only exit at the financial rand rate. Even after the abolition of the dual

exchange rate system in 1993, and with it the removal of controls on non-

residents, non-resident investors still were inadvertently caught inside the

exchange control net, and arguably still are (see, for example, National Treasury

2011)..

More concentrated industry. In 1985, one company, Anglo American,

controlled just over half of the Johannesburg Stock Exchange (see Table X), and

even in 2002, in about half of the main industrial products, the largest four firms

accounted for more than half of output (Chabane, Machaka, Molaba, Roberts, &

Taka, 2003), or, put another way, a quarter of production was controlled by four

firms.

Exchange controls did entirely create this problem – as noted above, the high

tariff barriers and lack of international competition due to sanction certainly

played a part, but trapping domestic capital within South Africa encouraged

firms to invest in non-core activities. This created huge conglomerates, with

interests in multiple industries.

Table 6 Largest conglomerates in 1985

1985 1990 1994 2002

Anglo American Corp 53.6 44.2 43.3 25.4

Sanlam 12.2 13.2 10.5 4.2

Liberty Life 2 2.6 7.2 3.2

Rembrandt / Remgro 3.8 13.6 13 13.7

SA Mutual / Old Mutual 10.6 10.2 9.7 13.3

Anglovaal 2.1 2.5 3.6

Four largest 80.2 81.2 76.5 56.6

Source: Adapted from (Chabane, Machaka, Molaba, Roberts, & Taka, 2003)

21

Reduced growth in export-oriented manufacturing. The stringent system of

exchange controls, and accompanying trade boycott may have been a

complicating factor in stimulating an export-oriented manufacturing sector.

Indeed, the 1980s were a “lost decade” for industry, with manufacturing growth

and, crucially for a labour-surplus economy, employment growth substantially

curtailed (McCarthy, 1994).

Increased exchange rate volatility for foreign investors. By creating a different

regime for resident and non-residents to trade in rand, the consequence may have

been to discourage a proper functioning foreign exchange market.

Slow development of a fully functioning forward market. At the very least, by

1993, it was clear that the foreign exchange forward market was substantially

underdeveloped (Ramos, 1991), and it has been the stated objective of successive

exchange control liberalisations to assist in develop this market.13

Lack of diversification for pension funds and other institutional investors.

The controls on residents extended to institutional funds, such as pension funds.

These were forced to invest only in South African assets, making it impossible for

trustees to diversify risk. Moreover, given the high concentration of mining shares

on the JSE, pension fund assets became highly correlated with the resources

cycle14

.

Captive investment base. The flip side of the diversification is that large South

African corporates were guaranteed a captive source of capital, and so did not

13

The widening of the net open forward position following the Asian crisis was, in part, due to the

Reserve Bank carrying losses from the state-guaranteed forward market. 14

The abolition of this control was one of the earliest after apartheid. In 1995, institutions were

able to use an asset swap mechanism to invest in offshore assets (Farrell & Todani, 2004). Since

then, the restriction has been further liberalised. The most recent was the decision in December

2011 to allow pension funds to have offshore assets up to 25 per cent of their total assets, with

higher limits for other institutional investors. This is no longer an “exchange control”, but rather a

macroprudential limit aimed at limiting pension fund exposure to foreign risk (e.g. currency and

sovereign risk). The limit is also variable, which provides some scope for reduction in the event of

a major exogenous shock.

22

compete with international firms for capital, creating internationally

uncompetitive South African corporates15

.

Reduction of the tax base. The figures quoted above, and the proceeds from the

voluntary disclosure programme, highlight that wealthy South African residents

and some companies were quite capable of exiting capital from South Africa, and

investing it offshore. Income from this capital was almost certainly not declared

in South Africa and essentially existed outside the tax net.

Some possible positives can also be identified. The onerous controls introduced in

1985 were designed to disincentivise foreign investors from wholesale

disinvestment, and may have delayed the implosion of the system. It is

questionable whether it the short-term benefit of stopping disinvestment

outweighed the medium- to long-term impact of an increasingly distorted and

isolated economy.

5 Conclusion

Exchange controls were part of a complex system of maintaining some financial

stability during apartheid, particularly as the system faced increased sanctions and

deteriorating external and internal conditions. However, the medium- to long-term

impact of the controls was potentially substantially negative, encouraging the

increased isolation of the apartheid state. The attempt to close off the economy to

protect the state may have had the perverse consequence of leaving the economy

structurally uncompetitive, creating a fragmented foreign exchange market, and

leaving the economy more fragile when democracy came in 1994. This legacy

lasted a number of years into the new dispensation.

15

In addition, the high tariff barriers to maintain the balance of payments surplus had arguably the

same result. There is a relatively large strand of literature (see for example Weeks or Sender

Standing and Weeks ) that argues that the post-apartheid economic liberalisation was

inappropriate, particularly following the Growth, Employment and Redistribution policy

announced in 1996. A detailed discussion of the effects of the post-Apartheid liberalisation period

is, however, outside the scope of this paper.

23

For policymakers today, many remnants of the exchange control system remain.

Understanding the context of why the particular parts of the system were

introduced as they were helps design future policy. The design of the future relies

on us understanding the past.

24

6 Bibliography

Barr, G., & Kantor, B. (1983). A theory of the financial rand discount. Retrieved

February 7, 2013, from Investment Analysts Society of South Africa:

www.iassa.co.za/articles/021_jun1983_01.pdf

Chabane, N., Machaka, J., Molaba, N., Roberts, S., & Taka, M. (2003). 10 year

review: Industrial structure and competition policy. Johannesburg: School

of Economic and Business Sciences, University of the Witwatersrand.

De Kock, G. (1954). A history of the South African Reserve Bank, 1920-1952.

Pretoria: J. L. van Schaik.

Farrell, G., & Todani, K. (2004). Capital flows, exchange control regulations and

exchange rate policy: The South African experience. OECD Seminar

"How to reduce debt costs in Southern Africa?". OECD.

Flood, R., & Marion, N. (1982). The Transmission of Disturbances under

Alternative Exchange-Rate Regimes with Optimal Indexing. Quarterly

Journal of Economics, 97(1), 43-66.

Garner, J. (1994). An analysis of the financial rand mechanism. LSE Centre for

Research into Economics and Finance in South Africa, Research paper no

9.

Garner, J., & Leape, J. (1991). South Africa's borrowings on international capital

markets: Recent developments in historical perspective. Research Paper

no 5. London: LSE: Centre for the Study of the South African Economiy

and International Finance.

25

Jones, S. (1994). South Africa's external trade in the 1980s. South African Journal

of Economic History Volume 9, No 2, September.

Kahn, B. (1991). Capital flight and exchange control in South Africa. Centre for

the Study of the South African Economy, Research paper 4. London

School of Economics.

Kahn, B. (1991). Capital flight and exchange controls in South Africa. LSE

Centre for the Study of the South African Economy and International

Finance, Research paper no 4.

Kiguel, M., & O'Connell, S. (1995). Parallel exchange rates in developing

countries. World Bank Research Observer, 10(1), February.

Leape, J. (1991). South Africa's foreign debt and the standstill. LSE Centre for the

Study of the South African Economy and International Finance, Research

paper 1.

McCarthy, C. (1994). The 1980s: a lost decade for the South African

manufacturing system. South African Journal of Economic History

Volume 9, No 2, September, 66-83.

Ovenden, K., & Cole, T. (1989). Apartheid and International Finance. Sydney:

Penguin.

Ramos, M. (1991). The development of the forward exchange market in South

Africa. LSE Centre for the Study of the South African Economy and

International Finance, Research paper 3.

Spandau, A. (1979). Economic Boycott Against South Africa. Cape Town: Juta.

Waldmeir, P. (1999). Anatomy of a Miracle: The end of apartheid and the birth of

the new South Africa. New York: W. W. Norton & Co.

26